Information processing system, information processing method, and program

Abstract

To indicate the health condition of an individual with high accuracy.SOLUTION: An information processing system includes an event information acquisition unit that acquires event information of a predetermined event type related to a living situation of a target person, the event information being related to an occurrence timing of an event of the event type, a data distribution generation unit that generates a data distribution representing an occurrence timing for each predetermined time window smaller than a predetermined time scale in the predetermined time scale on the basis of the event information, a regularity index data generation unit that generates regularity index data indicating regularity of the event type expressible by the occurrence timing for each predetermined time window on the basis of the data distribution, and an output unit that outputs the regularity index data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing system, an information processing method, and a program. [Background technology]

[0002] Conventionally, techniques for calculating scores related to an individual's health condition have been known.

[0003] For example, a lifestyle scoring system described in Patent Document 1 determines a physical activity score, which is the output value of a piecewise continuous physical activity function, and a sleep score, which is the output value of a piecewise continuous sleep function. The system also determines multiple nutrient scores, which are the output values ​​of a piecewise continuous nutrient function corresponding to each of multiple nutrients, and determines an energy score, which is the output value of a piecewise continuous energy function. The system then determines a dietary intake score by multiplying the average of each of the multiple nutrient scores by the energy score. The system then determines the lifestyle score by calculating the sum of (1) the physical activity score multiplied by a first weight, (2) the sleep score multiplied by a second weight, and (3) the dietary intake score multiplied by a third weight, and displays a representation of the lifestyle score on the display device. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Special table 2023-506385 publication Summary of the Invention [Problem to be solved by the invention]

[0005] However, an individual's health condition cannot be evaluated solely by the sum of the scores for each item, and can vary significantly depending on, for example, the consistency of lifestyle habits or the harmony of one's overall life. However, the lifestyle scoring system described in Patent Document 1 does not take into account at least the consistency of lifestyle habits or the harmony of one's overall life, and therefore cannot accurately indicate an individual's health condition.

[0006] Therefore, an object of the present invention is to provide a highly accurate indication of an individual's health condition. [Means for solving the problem]

[0007] An information processing system according to one aspect of the present invention includes an event information acquisition unit that acquires event information of a predetermined event type related to the living situation of a subject, the event information being event information related to the occurrence timing of events of the event type; a data distribution generation unit that generates, based on the event information, a data distribution that represents the occurrence timing for each predetermined time window on a predetermined time scale that is smaller than the predetermined time scale; a regularity index data generation unit that generates, based on the data distribution, regularity index data that indicates a regularity of the event type that can be expressed by the occurrence timing for each predetermined time window; and an output unit that outputs the regularity index data. [Effects of the Invention]

[0008] According to the present invention, the health condition of an individual can be indicated with high accuracy. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram showing an overview of processing in a score processing system 100 according to an embodiment of the present invention. [Figure 2] 1 is a diagram showing the configuration of a score processing system 100 according to an embodiment of the present invention. [Figure 3] 10 is a diagram showing an example of event information stored in a storage unit 110. FIG. [Figure 4] FIG. 2 is a schematic diagram showing a weighting matrix W. [Figure 5A] 10 is a diagram showing an example of a regularity index displayed by the display unit 155. FIG. [Figure 5B] 10 is a diagram showing an example of a regularity index displayed by the display unit 155. FIG. [Figure 6] FIG. 10 is a diagram showing an example of a total score displayed by the display unit 155. [Figure 7] 10 is a flowchart showing an example of processing in the score processing system 100. [Figure 8] 10 is a flowchart showing an example of processing in the score processing system 100. [Figure 9] FIG. 9 is a diagram illustrating an example of the hardware configuration of a computer 900. DETAILED DESCRIPTION OF THE INVENTION

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will now be described with reference to the accompanying drawings, in which: Figure 1 is a diagram showing an overview of processing in a score processing system 100 according to one embodiment of the present invention.

[0011] The score processing system 100 is an information processing system that executes the first calculation process and the second calculation process.

[0012] In the first calculation process, the score processing system 100 calculates a regularity index. In this case, the score processing system 100 is executed by a program that calculates a regularity index.

[0013] Specifically, first, the score processing system 100 acquires event information of a predetermined event type related to the living situation of the subject, which is event information related to the occurrence timing of events of the event type. Next, the score processing system 100 generates regularity index data indicating regularity regarding the occurrence timing for each predetermined time window on a predetermined time scale based on a data distribution regarding the occurrence timing based on the event information (S101). At this time, the score processing system 100 may generate regularity index data according to at least one of the kernel regularity index, KL divergence, Wasserstein distance, and circular Wasserstein distance, which will be described later, for example.

[0014] In the second calculation process, the score processing system 100 calculates an overall score (HAI) indicating the health status of the subject based on the item scores of multiple items. In this case, the score processing system 100 is executed by a program that calculates the overall score based on the item scores.

[0015] Specifically, first, the score processing system 100 calculates scores for each of a plurality of items related to the living situation of the subject based on a regularity score (S102) calculated based on data distribution related to regularity and information indicating questionnaire results, etc. (S103). Next, the score processing system 100 calculates a harmony score indicating the harmony of the scores for each of the plurality of items based on the scores for each of the plurality of items and interaction information regarding interactions between the plurality of items (S104). Then, the score processing system 100 calculates a total score indicating the health situation of the subject based on the scores for each of the plurality of items and the harmony score (S105).

[0016] 2 is a diagram showing the configuration of a score processing system 100 according to one embodiment of the present invention. The score processing system 100 is communicably connected to a subject device 200 via a network such as the Internet. Details of the score processing system 100 will be described later.

[0017] Subject device 200 is an information processing device used by a subject for whom scores (regularity score, harmony score, and overall score) described below are to be calculated, and is, for example, a computer, a smartphone, a tablet terminal, a personal computer, or the like.

[0018] The subject accesses the score processing system 100 using the subject device 200. The subject device 200 provides the score processing system 100 with, for example, information indicating the results of a questionnaire given to the subject (described later), information on biomarkers based on genetic information, and information acquired from a wearable device. The subject device 200 can also acquire, for example, scores (regularity score, harmony score, and overall score) output from the score processing system 100 and can be displayed by a display unit 155 (described later). The information on biomarkers and information acquired from the wearable device may be transmitted from a measurement device to the score processing system 100 along with the subject's identification information. The score processing system 100 may acquire, for example, medical interview data, physical measurement data such as height and weight as information indicating the questionnaire results, images such as electrocardiograms, arm strength, CT scans, MRI scans, etc. as information on biomarkers, and information such as pulse rate and heart rate as information acquired from the wearable device.

[0019] Although one subject device 200 is shown in FIG. 2, a plurality of subject devices 200 may be provided.

[0020] Next, details of the score processing system 100 will be described. The score processing system 100 includes a storage unit 110, an acquisition unit 115, a setting unit 120, a data distribution generation unit 123, a regularity index data generation unit 125, a regularity score calculation unit 130, an item score calculation unit 135, a harmony score calculation unit 140, a total score calculation unit 145, an age prediction unit 148, an output unit 150, and a display unit 155. Each unit shown in Fig. 2 can be realized, for example, by using a storage area or by a processor executing a program stored in the storage area.

[0021] The storage unit 110 stores information to be processed in the score processing system 100 .

[0022] The acquisition unit 115 acquires information indicating the results of a questionnaire given to the subject regarding a plurality of items, information regarding biomarkers based on genetic information, and information acquired from the wearable device, and stores the information in the storage unit 110. This information is, for example, information used by the item score calculation unit 135 (described later) when calculating item scores. This allows the score processing system 100 to use item scores calculated within the score processing system 100.

[0023] Here, the multiple items may be items related to "living environment status" (Environment), "Body Function status" (Body Function), "Stress status" (QOL), "Nutrition status" (Nutrition), "Activity status" (Activity), "Sleep status" (Sleep), and "Personal" (Personal). "Personal" is, for example, information about the individual subject.

[0024] Furthermore, the acquisition unit 115 acquires the scores of each of a plurality of items related to the living situation of the subject, and stores item score information related to the acquired item scores in the storage unit 110. At this time, the acquisition unit 115 may acquire, for example, the item scores calculated by the item score calculation unit 135. This allows the score processing system 100 to calculate the item scores within the score processing system 100.

[0025] The acquiring unit 115 may also acquire item scores calculated in an external information processing system. This allows the score processing system 100 to use item scores calculated outside the score processing system 100.

[0026] Furthermore, the acquisition unit 115 (for example, the event information acquisition unit) acquires event information of a predetermined event type related to the living situation of the subject, which is event information related to the occurrence timing of an event of the event type, and stores the acquired event information in the storage unit 110. The occurrence timing of the event includes, for example, the time when the event occurred.

[0027] Here, the event type includes, for example, an event type capable of measuring lifestyle habits, particularly, for example, at least one of the subject's activity status, sleep status, and nutritional status. The event type may include multiple event types. For example, the occurrence timing of the activity status may include the time when exercise started and ended, the occurrence timing of the sleep status may include the time when sleep started (time when sleep fell asleep) and the time when wake-up (time when sleep ended), and the nutritional status may include the time when nutrition intake started and ended.

[0028] Furthermore, to go into more detail about nutritional status, it is best to judge nutritional status from two main perspectives. The first is balance. A diet of appropriate quantity and quality is essential for maintaining and improving health, so the perspective is whether a person is eating a nutritionally balanced diet. This can be an assessment of dietary balance, such as staple foods, main dishes, side dishes, dairy products, and fruit; nutritional components such as carbohydrates, proteins, and lipids; or estimated calories calculated using images. The other perspective is whether the person is eating regularly. In addition, if there is data on nutritional status from blood tests, values ​​such as total protein, albumin, red blood cell count, blood sugar level, lipid-related data, and electrolyte data may be looked at, and even if there is no test data, fluctuations in BMI and weight may be looked at.

[0029] Regarding balance, for example, the user is asked to input a nutritional balance sheet and beverage intake, and a judgment is made based on these to determine whether the diet is balanced. Here, a nutritional balance sheet is a sheet that confirms whether a meal is balanced in terms of staple foods, main dishes, side dishes, dairy products, fruit, etc. Furthermore, beverage intake includes, for example, alcohol intake, caffeine intake, and sugar intake. When judging the level of balance, the judgment can be made based on either the nutritional balance sheet or beverage intake alone, or a combination of both.

[0030] On the other hand, with regard to regular eating behavior, the user is asked to input the timing of meal times, the frequency of meals (lunch, lunch, dinner), the frequency of eating and drinking, etc., and the level of regularity regarding regular eating behavior is judged based on such eating behavior. In this judgment, the exemplified eating behaviors may be judged based on each individual eating behavior, or a judgment may be made based on a combination of multiple eating behaviors.

[0031] The regularity of the nutritional status is judged using a predetermined standard based on balanced and regular eating habits (a score is calculated using a predetermined calculation formula). For example, even if a person has regular eating habits, if their eating habits are unbalanced, their nutritional status may be judged to be poor.

[0032] Furthermore, the acquiring unit 115 can acquire interaction information regarding interactions between a plurality of items or a weight matrix W, which will be described later, and store it in the storage unit 110. The interaction information or the weight matrix W may be information set by an administrator of the score processing system 100, for example. That is, the acquiring unit 115 acquires the interaction information or the weight matrix W based on an operation (for example, an input operation) of the administrator. Details of the interaction information and the weight matrix W will be described later.

[0033] The setting unit 120 sets the time scale for calculating the regularity index, and stores the set time scale setting information in the storage unit 110. Here, the time scale may be, but is not limited to, weekly, monthly, or yearly.

[0034] The data distribution generating unit 123 generates distribution information indicating a data distribution of the occurrence times of events according to the event type, for example, as the event information acquired by the acquiring unit 115. Specifically, when the event type is a sleep state, the data distribution indicating the occurrence times of events may be a distribution of sleep onset time, sleep end time, sleep duration, and sleep duration for each sleep stage (e.g., awake, REM sleep, light sleep, deep sleep, etc.).

[0035] FIG. 3 is a diagram showing an example of data distribution information stored in the storage unit 110. The data distribution stored in the storage unit 110 includes, for example, a subject ID and distribution information. The distribution information may be, for example, distribution information every 5 minutes as shown in FIG. 3, but the distribution information is not limited to this and may be, for example, distribution information every hour. Furthermore, the distribution information may include, for example, distribution information for each day over multiple days. Furthermore, the distribution information may be aggregating occurrence times according to events (e.g., sleep and falling asleep) for each predetermined time window.

[0036] The regularity index data generation unit 125 generates regularity index data based on the data distribution regarding the occurrence timing of events based on the event information, which is generated by the data distribution generation unit 123. The regularity index data is data indicating the regularity of an event type, which can be expressed, for example, by the occurrence timing of event information for each predetermined time window on a predetermined time scale that is smaller than the predetermined time scale. The regularity index data generation unit 125 stores the generated regularity index data in the storage unit 110.

[0037] Here, regularity index data is a new concept in public health that indicates the consistency of a person's activities. Specifically, it is an index that indicates the consistency of measurable lifestyle events over a certain time scale (e.g., consistency in the timing of events), and indicates the degree of regularity of a subject's lifestyle. The regularity index may also be a multidimensional index of regularity. The multidimensional regularity index (Multi Model Regularity Index) provides a new method for calculating the regularity index of events and can be used for other possible lifestyle events besides sleep.

[0038] The regularity index data is, for example, data indicating the occurrence frequency or occurrence probability of a predetermined event (e.g., sleep) for each predetermined time window that is smaller than a predetermined time scale. The regularity index data can indicate the regularity of a person's activity by the change in the occurrence frequency or occurrence probability of the event for each time window. For example, if the range of the time window in which the change in occurrence frequency or occurrence probability occurs is within a predetermined time, it indicates that the activity is being performed regularly, and if it exceeds the predetermined time, it indicates that the event is occurring beyond the predetermined time, indicating that the activity is being performed irregularly.

[0039] The regularity index data generating unit 125 can generate regularity index data that can display figures corresponding to values ​​based on each of the plurality of individual regularity index data for each predetermined time window that is smaller than a predetermined time scale, for example, arranged in a circular shape, based on the plurality of individual regularity index data.

[0040] The regularity index data generating unit 125 can generate regularity index data for the set time scale based on the time scale setting information. For example, the regularity index data generating unit 125 aggregates data according to the set time scale to generate a data distribution, and generates regularity index data based on the generated data distribution.

[0041] Furthermore, the regularity index data generating unit 125 can calculate a plurality of individual regularity index data corresponding to each of the plurality of event types. For example, the regularity index data generating unit 125 may generate the above-mentioned regularity index data (individual regularity index data) for each of the activity status, sleep status, and nutritional status included in the plurality of event types.

[0042] The regularity score calculation unit 130 calculates regularity scores related to regularity in the living situation of the subject, and stores the calculated item scores in the storage unit 110. Here, the living situation includes at least one of the stress situation, physical function situation, living environment situation, activity situation, sleep situation, and nutritional situation of the subject.

[0043] The regularity score calculation unit 130 can calculate a regularity score, which is, for example, a value between 1 and 100, based on the data distribution or regularity index data generated by the data distribution generation unit 123. As an example, the regularity score calculation unit 130 may calculate a regularity score indicating the regularity of an event using the variance of a time window indicating the occurrence of an event, based on distribution information indicating the data distribution of the occurrence times of events according to the event type. For example, if the variance of the time window indicating the occurrence of an event is large, the score will be small, and if the variance of the time window is small, the score will be large.

[0044] The regularity score, for example, represents the status of measurable events in a person's daily lifestyle on a scale of 1 to 100. A regularity score of 100, for example, means that the person has a high degree of consistency in their measurable events, and a regularity score of 1, for example, means that the person has no consistency at all.

[0045] The regularity score calculation unit 130 may calculate the regularity score based on the degree of deviation between a data distribution regarding the timing of occurrence of an event in the event information and a reference data distribution, which is a reference data distribution. For example, the reference data distribution may indicate a data distribution when an ideally regular activity is performed. In this case, a higher deviation from the reference data distribution indicates more irregularity, and a lower deviation indicates more regularity. Furthermore, the reference data distribution may indicate a data distribution when the most irregular activity (e.g., random activity) is performed. In this case, a higher deviation from the reference data distribution indicates more regularity, and a lower deviation indicates more irregularity.

[0046] A specific example of the process of calculating the regularity score by the regularity score calculation unit 130 will be described. When calculating the regularity score, the regularity score calculation unit 130 uses, for example, a data distribution for the subject based on the event information generated by the data distribution generation unit 123. Specifically, first, the data distribution generation unit 123 assigns, for example, an event indicated by the event information to each time window. At this time, the data distribution generation unit 123 assigns, for example, the event to a time window that is closest to or includes the occurrence time of the event, thereby obtaining a distribution of the event. Next, the data distribution generation unit 123 normalizes the distribution of the event to obtain a discrete probability distribution P(X). Here, the discrete probability distribution is an empirical distribution that indicates the frequency or probability that an event will be executed in a predetermined time window.

[0047] The regularity score calculation unit 130 then calculates the regularity score based on, for example, the degree of deviation between a data distribution (e.g., discrete probability distribution P(X)) relating to the occurrence timing of an event in the event information and a reference data distribution, which is a reference data distribution. In this case, the regularity score calculation unit 130 can calculate the regularity score based on, for example, the degree of deviation between the data distribution and the reference data distribution according to at least one of KL divergence, Wasserstein distance, and circular Wasserstein distance. Note that the data distribution (e.g., discrete probability distribution P(X)) relating to the occurrence timing of an event in the event information may be regularity index data generated by the regularity index data generation unit 125.

[0048] The regularity score calculation unit 130 can calculate the regularity score based on KL divergence (Kullback-Leibler divergence) using the following formula.

[0049]

number

[0050] Here, for example, P represents a data distribution (discrete probability distribution) regarding the timing of occurrence of an event in the event information, U represents a reference distribution (e.g., a uniform distribution), X represents a predetermined time or time window, E represents the length of the period, B represents the length of the time window, and N represents the number of periods.

[0051] Furthermore, the regularity score calculation unit 130 can calculate the regularity score based on the Wasserstein distance using the following formula:

[0052]

number

[0053] Here, for example, μ indicates a data distribution relating to the occurrence timing of an event in the event information, and ν indicates a reference distribution (for example, a uniform distribution).

[0054] Furthermore, the regularity score calculation unit 130 can calculate the regularity score based on the circular Wasserstein distance using the following formula:

[0055]

number

[0056] Here, for example, μ indicates a data distribution relating to the occurrence timing of an event in the event information, and ν indicates a reference distribution (for example, a uniform distribution).

[0057] Furthermore, the regularity score calculation unit 130 may calculate the regularity score based on a data distribution (e.g., a discrete probability distribution P(X)) regarding the occurrence timing of an event in the event information. In this case, the regularity score calculation unit 130 can calculate, for each predetermined time window smaller than a predetermined time scale, a regularity score according to a kernel regularity index (KRI) based on the similarity of the content of events on consecutive predetermined time scales. Here, the KRI is calculated, for example, by the following formula:

[0058]

number

[0059] If the predetermined time scale is one day, s_i is the string representation of the event on the i-th day, and s_(i+1) is the string representation of the event on the next day. The string representation may be, for example, "1" if asleep and "0" if awake. Furthermore, k indicates a time window (for example, 5 minutes or 1 hour). D is the total number of days recorded, and δ is a predetermined function for scoring whether the event records of two consecutive days are similar with a score between 0 and 1. δ may be, for example, a function that outputs "1" if the string representations of the two consecutive days match, and "0" if they do not match.

[0060] The calculation process for a regularity score based on a KRI has at least three features. The first feature is that the KRI can calculate more accurate regularity index data based on the data distribution for a specific time period within a day, including multiple specific time windows k, rather than the entire day. For example, if the regularity score for sleep is calculated by taking into account the data distribution during the daytime, when people are awake more frequently, an excessively high (i.e., regular) index may be calculated. The KRI can avoid this problem. The second feature is that a flexible scoring function δ can be used, making it possible to identify individuals based on age and gender. The third feature is that the regularity score can be calculated based on an arbitrarily configurable time window k.

[0061] The item score calculation unit 135 calculates an item score for each of the multiple items, and stores the calculated item scores in the storage unit 110. The item score is a value between 1 and 100, for example.

[0062] The item score calculation unit 135 can calculate the item score for each item based on, for example, information indicating the results of a questionnaire given to the subject, information on biomarkers based on genetic information, and information acquired from a wearable device.

[0063] In this case, the item score calculation unit 135 can calculate the item score of each item as the sum of all elements of the product of score_i and weight_i for each element i, based on the score (score_i) of element i in the item and the weight (weight_i) of the element. Here, the element may be, for example, an element associated with a question in a questionnaire for the subject, one biomarker based on genetic information, and one piece of information obtainable from a wearable device.

[0064] Furthermore, the item score calculation unit 135 can calculate the item score further based on the regularity score. For example, the item score calculation unit 135 may calculate the item score by performing a predetermined calculation on the regularity score and a value calculated as the sum of all elements of the product of score_i and weight_i for each element i. Here, the predetermined calculation may be, for example, a summation, or may be a summation of the value calculated as the sum of all elements of the product of score_i and weight_i for each element i and a value obtained by multiplying the regularity score by a predetermined coefficient (weight).

[0065] Furthermore, the item scores calculated based on the regularity score may be the item scores of all of the multiple items, or may be the item scores of some of the multiple items (e.g., nutritional status, activity status, sleep status).

[0066] The harmony score calculation unit 140 calculates a harmony score indicating the harmony of each score of the multiple items based on the scores of each of the multiple items and interaction information regarding the interactions between the multiple items, and stores the calculated harmony score in the memory unit 110.

[0067] Here, the harmony score is, for example, a score indicating how well each item score of a plurality of items in a subject is in harmony with the interaction network indicated by the interaction information. That is, for example, the closer the relationship between the item scores is to the interaction network, the higher the harmony score will be, and the less the relationship between the item scores is to the interaction network, the lower the harmony score will be. The harmony score is, for example, a value between 0 and 1.

[0068] The harmony score calculation unit 140 can calculate the harmony score based on, for example, a Rayleigh quotient of a vector indicating the scores of each of the multiple items and a matrix indicating the interaction information. A specific example of the harmony score calculation process will be described.

[0069] The vector s indicating the scores of each of the multiple items is, for example, a vector in which the item scores of each item are arranged. If there are seven items, the vector s is a seven-dimensional vector.

[0070] The matrix L indicating the interaction network is calculated as follows using a 7×7 weight matrix W indicating the weights (values ​​of 0 to 1) between items and a diagonal matrix D.

[0071]

number

[0072] FIG. 4 is a schematic diagram showing an interaction network for a weight matrix W. For example, the weight from the sleep situation to the activity situation is 0.3, and the weight from the activity situation to the sleep situation is 0.15. In this case, in the weight matrix W, for example, the value at the intersection of the row of the sleep situation and the column of the activity situation is 0.3, and the value at the intersection of the row of the activity situation and the column of the sleep situation is 0.15. The diagonal matrix D is a diagonal matrix in which the diagonal components represent the weights of each item. Note that the weight matrix W and the diagonal matrix D can be set arbitrarily.

[0073] Then, the harmony score calculation unit 140 calculates the Rayleigh quotient for the vector s and the matrix L as the harmony score h using the following formula:

[0074]

number

[0075] The total score calculation unit 145 calculates a total score indicating the health condition of the subject based on the harmony score, and stores the calculated total score in the storage unit 110.

[0076] Furthermore, the overall score calculation unit 145 can calculate the overall score further based on the item scores. At this time, the overall score calculation unit 145 can calculate the overall score based on the item scores calculated based on the regularity scores. Here, the overall score may be called a Healthy Aging Index (HAI).

[0077] The overall score calculation unit 145 can calculate the overall score HAI, for example, using the following formula:

[0078]

number

[0079] where h denotes the agreeableness score, s denotes the item score, and s A ~s G indicates the item score for each of the seven items.

[0080] Furthermore, the overall score calculation unit 145 can calculate the contribution of each of the item score and the harmony score to the overall score based on SHAP (SHapley Additive exPlanations). At this time, the overall score calculation unit 145 can calculate the contribution of each of the item score and the harmony score to the overall score, for example, by the following formula:

[0081]

number

[0082] where S is the set of all subsets of item scores and agreement scores. The contribution can be calculated using existing techniques available to those skilled in the art.

[0083] The age prediction unit 148 performs machine learning using each score Sn (n=A to G) used in calculating the above-mentioned overall score HAI and the measured blood biomarkers. The machine learning may be unsupervised learning without labels, or may be supervised learning using the user's age as the correct label. Here, it is assumed that unsupervised learning is performed.

[0084] This machine learning predicted biological age extends the healthy aging index by using blood biomarker information and machine learning-based risk prediction. The machine learning model is trained on a large dataset to make a good prediction of a person's biological age. A negative gap between the predicted biological age and chronological age indicates good health and a low risk of mortality.

[0085] An example of an algorithm for calculating biological age is as follows: 1. Calculate the seven HAI index scores. 2. Aggregate and process index level data from blood biomarkers. 3. The data from 1. and 2. are used as input for a pre-trained machine learning model for risk prediction. 4. Perform a forward pass to obtain predicted biological age and associated mortality risk score.

[0086] An example of a learning algorithm for a machine learning model is as follows: 1. Build a machine learning architecture. 2. Prepare the dataset (blood biomarker information and each HAI score Sn). 3. Unsupervised learning on training data (unlabeled). 4. Learning with labeled cases. 5. Score the model for future predictions.

[0087] By performing the above process, the age prediction unit 148 predicts the biological age based on each HAI score Sn and blood biomarker information. For example, if a user has a predicted biological age of 55 but is actually 60 years old, it can be said that this user has a lower risk of death than the average 60-year-old person.

[0088] The output unit 150 outputs at least one of the regularity index data, the regularity score, the item score, the harmony score, the overall score, and the risk based on the predicted biological age. The output unit 150 can output at least one of the regularity index data, the regularity score, the item score, the harmony score, the overall score, and the risk based on the predicted biological age to, for example, the subject device 200. For example, the output unit 150 outputs at least one of the regularity index data, the regularity score, the item score, the harmony score, the overall score, and the risk based on the predicted biological age to be displayed on the display unit 155. Note that the output unit 150 may output at least one of the regularity index data, the regularity score, the item score, the harmony score, the overall score, and the risk based on the predicted biological age to be displayed as a figure, graph, or the like.

[0089] The display unit 155 displays the results of the processing in the score processing system 100. The display unit 155 can display the results of the processing in the score processing system 100 on the display screen of the subject device 200, for example.

[0090] 5A and 5B are diagrams showing an example of a display in which regularity index data is visualized by the display unit 155. FIG.

[0091] Based on the individual regularity index data, the display unit 155 displays, for example, a circular arrangement of figures corresponding to values ​​based on the individual regularity index data for each predetermined time window smaller than a predetermined time scale. In FIG. 5A , for example, each bar-shaped figure corresponds to a value based on each of the individual regularity indexes. One circle represents one day, and a bar-shaped figure corresponding to the individual regularity index for each time window is displayed. The length of the bar-shaped figure represents, for example, the cumulative value of the occurrence probability of an event. The longer the bar length, the higher the occurrence probability. Specifically, for example, the longest bar-shaped figure that can be displayed on the screen represents a 100% occurrence probability. According to this example, if the sleep onset time is constant, the occurrence probability increases within the same time window, and the bar-shaped figure for the time window including the sleep onset time suddenly becomes longer in successive bar figures. On the other hand, if the sleep onset times are dispersed, the time windows with occurrence probabilities are dispersed, and the bar-shaped figure for the time window including each sleep onset time gradually becomes longer in successive bar figures. That is, in the change in the length of the continuous bar-shaped figures, if the length suddenly increases, it indicates that the subject is falling asleep regularly, and if the length gradually increases, it indicates that the subject is falling asleep irregularly.

[0092] The display unit 155 may display, for example, individual regularity index data corresponding to one item on a single circumference, or may display multiple individual regularity index data on a single circumference. When the display unit 155 displays multiple individual regularity index data on a single circumference, for example, the multiple individual regularity index data may each be displayed in a different color, pattern, or shape, such that the individual regularity index data for a predetermined item is displayed within a predetermined time window. The predetermined time window within which the individual regularity index data for a predetermined item is displayed may be set, for example, by a person referring to the display (e.g., the subject) or an administrator of the score processing system 100. In FIG. 5A , for example, an area 510 displays individual regularity index data corresponding to a sleep situation. Furthermore, arcs 520 and 530 displayed in the portions of the area 510 shown in FIG. 5A where the length of the bars changes indicate regularity, with shorter arcs indicating more regularity and longer arcs indicating more irregularity. Furthermore, an area other than the area 510 may display individual regularity index data corresponding to, for example, an activity situation. This is because sleep status and activity status are mutually exclusive events that do not occur simultaneously.

[0093] FIG. 5B is an example of another graphic displayed on the display unit 155. In FIG. 5B, for example, a curve plotting distance from the center corresponds to a value based on a certain individual regularity index. As in FIG. 5A, one cycle represents one day, and curves corresponding to the individual regularity indexes of each time window are displayed. The curves represent, for example, the cumulative value of the occurrence probability of an event, with the greater the distance from the center, the higher the occurrence probability. Specifically, for example, the portion of the curve that is the farthest from the center that can be displayed on the screen represents a 100% occurrence probability. According to FIG. 5B, if the sleep onset time is constant, the occurrence probability increases within the same time window, and the angle of the curved portion of the time window including the sleep onset time becomes smaller (e.g., W2) in the curved diagram plotting the distance from the center. The angle W indicates the angle from 0% to 100% for the cumulative value of the occurrence probability. On the other hand, if the sleep onset times are dispersed, the time windows with occurrence probabilities will be dispersed, and the angle of the curved portion of the time window including each sleep onset time will become larger (e.g., W1) in the curved diagram plotting the distance from the center. That is, in the change in the curve plotting the distance from the center, a small angle indicates that the person is falling asleep regularly, and a large angle indicates that the person is falling asleep irregularly.

[0094] As in the case of FIG. 5A, in the case of FIG. 5B, the display unit 155 may display, for example, individual regularity index data corresponding to one item on a single circumference, or may display multiple individual regularity index data on a single circumference. When the display unit 155 displays multiple individual regularity index data on a single circumference, for example, the multiple individual regularity index data may each be displayed in a different color, pattern, or shape, so that the individual regularity index data for a predetermined item is displayed within a predetermined time window. The predetermined time window during which the individual regularity index data for a predetermined item is displayed may be set, for example, by a person referencing the display (e.g., the subject) or an administrator of the score processing system 100. In FIG. 5B, the central angle of the curved portion plotted against the distance from the center indicates regularity, with a smaller central angle indicating more regularity and a larger central angle indicating more irregularity. Furthermore, in the curved portion plotted against the distance from the center, the area of ​​the curved portion with the greatest distance from the center may display individual regularity index data corresponding to, for example, an activity status. This is because sleep status and activity status are mutually exclusive events that do not occur simultaneously.

[0095] In addition, in Fig. 5B, the shaded area of ​​the curve shows individual regularity index data corresponding to the sleep situation. In the shaded area shown in Fig. 5B, the width of the rise of the curve from the inner circle indicating a 0% occurrence probability to the outer circle indicating a 100% occurrence probability indicates regularity, and it can be said that the narrower the width, the more regular (e.g., W2), and the wider the width, the more irregular (e.g., W1).

[0096] FIG. 6 is a diagram showing an example of the total score displayed by the display unit 155. As shown in FIG.

[0097] The display unit 155 displays a gauge corresponding to the overall score and multiple gauges corresponding to the item scores of multiple items, arranged, for example, radially. The item scores may be item scores based on the regularity score. The display example shown in FIG. 6 includes an area 610 displaying the overall score, an area 620 displaying a gauge corresponding to the overall score, and an area 630 displaying a gauge corresponding to the item score. Each score value may be displayed in a different color. A subject referring to the display in FIG. 6 can intuitively grasp the level of the overall score and the level of each item score.

[0098] 7 is a flowchart showing an example of processing in the score processing system 100. FIG. 7 is a flowchart showing an example of processing in the score processing system 100 for calculating regularity index data.

[0099] First, the acquiring unit 115 acquires event information of a predetermined event type related to the living situation of the subject, which is event information at a plurality of different times related to the occurrence timing of an event of the event type (S701). The acquiring unit 115 acquires, as the event information, for example, occurrence time information of an event according to the event type.

[0100] The setting unit 120 sets the time scale for generating regularity index data (S702). Note that the process performed by the setting unit 120 is not necessarily required.

[0101] The data distribution generating unit 123 generates, as the event information acquired by the acquiring unit 115, distribution information indicating the data distribution of the occurrence times of events according to the event types, for example (S703).

[0102] Next, the regularity index data generating unit 125 generates regularity index data indicating the regularity of the event type, which can be expressed, for example, by the occurrence timing of the event information for each predetermined time window on a predetermined time scale that is smaller than the predetermined time scale, based on the data distribution regarding the occurrence timing of the event information (S704).

[0103] The output unit 150 outputs the regularity index data, and the display unit 155 displays the regularity index data (S705). The display unit 155 displays, for example, figures corresponding to values ​​based on each of a plurality of individual regularity indexes for each predetermined time window smaller than a predetermined time scale, arranged, for example, in a circular shape (see, for example, FIG. 5).

[0104] 8 is a flowchart showing an example of processing in the score processing system 100. FIG. 8 is a flowchart showing an example of processing in the score processing system 100 to calculate a total score.

[0105] The acquiring unit 115 acquires information indicating the results of a questionnaire given to a subject, information on biomarkers based on genetic information, and information acquired from a wearable device (S801).

[0106] The regularity score calculation unit 130 calculates a regularity score having a value of, for example, 1 to 100 based on the regularity index data or the data distribution information (S802). The calculation process of the regularity index data may be, for example, a process based on the flowchart shown in FIG.

[0107] The item score calculation unit 135 calculates the item score based on the information indicating the questionnaire results for the subject, the information on the biomarkers based on the genetic information, the information acquired from the wearable device, and the regularity score (S803). At this time, the item score calculation unit 135 may calculate the item score for each item as the sum of all elements of the product of score_i and weight_i for each element i, for example, based on the score (score_i) of element i in the item and the weight (weight_i) of the element.

[0108] The harmony score calculation unit 140 calculates the harmony score based on the item scores (S804). At this time, the harmony score calculation unit 140 may calculate the harmony score based on, for example, a Rayleigh quotient of a vector indicating the scores of each of the multiple items and a matrix indicating the interaction information.

[0109] The overall score calculation unit 145 calculates an overall score based on the harmony score and the item scores (S805). The output unit 150 outputs the overall score, and the display unit 155 displays the overall score and the item scores (S806). At this time, the display unit 155 may display a gauge corresponding to the overall score and multiple gauges corresponding to the item scores of the multiple items, arranged, for example, radially.

[0110] Next, an example of a hardware configuration in which the score processing system 100 is realized by a computer 900 will be described with reference to Fig. 9. Fig. 9 is a diagram showing an example of the hardware configuration of the computer 900.

[0111] As shown in FIG. 9, a computer 900 includes, for example, a processor 901, a memory 902, a storage device 903, an input I / F unit 904, a data I / F unit 905, a communication I / F unit 906, and a display device 907.

[0112] Computer 900 may be, for example, a server computer, a personal computer (e.g., desktop, laptop, tablet, etc.), a media computing platform (e.g., cable, satellite set-top box, digital video recorder, etc.), a handheld computing device (e.g., PDA, email client, etc.), or any other type of computing or communications platform.

[0113] The processor 901 is a control unit that controls various processes in the computer 900 by executing programs stored in the memory 902 .

[0114] The memory 902 is a storage medium such as a RAM (Random Access Memory), etc. The memory 902 temporarily stores the program code of the program executed by the processor 901 and data required when the program is executed.

[0115] The storage device 903 is a non-volatile storage medium such as a hard disk drive (HDD), flash memory, etc. The storage device 903 stores an operating system and various programs for realizing the above-mentioned configurations.

[0116] The input I / F unit 904 is a device for receiving input from a user. The input I / F unit 904 is, for example, a keyboard, a mouse, a touch panel, various sensors, a wearable device, etc. The input I / F unit 904 may be connected to the computer 900 via an interface such as a USB (Universal Serial Bus).

[0117] The data I / F unit 905 is a device for inputting data from outside the computer 900. The data I / F unit 905 is, for example, a drive device for reading data stored in various storage media. The data I / F unit 905 may be provided outside the computer 900. When the data I / F unit 905 is provided outside the computer 900, the data I / F unit 905 is connected to the computer 900 via an interface such as a USB.

[0118] The communication I / F unit 906 is a device for performing data communication via a network such as the Internet, either wired or wirelessly, with devices external to the computer 900. The communication I / F unit 906 may be provided outside the computer 900. When the communication I / F unit 906 is provided outside the computer 900, the communication I / F unit 906 is connected to the computer 900 via an interface such as a USB.

[0119] The display device 907 is a device for displaying various types of information. The display device 907 is, for example, a liquid crystal display, an organic EL (Electro-Luminescence) display, a display of a wearable device, or the like. The display device 907 may be provided outside the computer 900. When the display device 907 is provided outside the computer 900, the display device 907 is connected to the computer 900 via, for example, a display cable. Furthermore, when a touch panel is adopted as the input I / F unit 904, the display device 907 may be configured as an integral part of the input I / F unit 904.

[0120] Each component of the score processing system 100 may be realized by a hardware configuration similar to that of the computer 900 shown in Fig. 9. For example, the processor 901 may function as the acquisition unit 115, setting unit 120, data distribution generation unit 123, regularity index data generation unit 125, regularity score calculation unit 130, item score calculation unit 135, harmony score calculation unit 140, total score calculation unit 145, age prediction unit 148, and output unit 150 of the score processing system 100 by executing a program that instructs a computer to perform each process of the above-described embodiments. The memory 902 and / or the storage device 903 may function as the storage unit 110 of the score processing system 100. The display device 907 may function as the display unit 155 of the score processing system 100.

[0121] The above describes one embodiment of the present invention. The score processing system 100 acquires scores for each of a plurality of items related to the living situation of a subject, calculates a harmony score based on the scores for each of the items and interaction information, and outputs the harmony score. In this way, the score processing system 100 can calculate a harmony score that accurately indicates the subject's health condition, taking into account the harmony of the subject's overall life.

[0122] Furthermore, the score processing system 100 can calculate a harmony score based on a Rayleigh quotient of a vector indicating the scores of each of a plurality of items and a matrix indicating interaction information, thereby enabling the score processing system 100 to calculate a harmony score that indicates the health condition of a subject with high accuracy.

[0123] Furthermore, the score processing system 100 can calculate a total score that indicates the health condition of the subject based on the harmony score. This allows the score processing system 100 to calculate a total score that indicates the health condition of the subject with high accuracy based on the harmony score.

[0124] Furthermore, the score processing system 100 can calculate a regularity score and calculate a total score based on the regularity score, thereby enabling the score processing system 100 to calculate a total score that accurately indicates the health condition of an individual, further taking into consideration the consistency of lifestyle habits.

[0125] Furthermore, the score processing system 100 can radially display a gauge corresponding to the total score and a plurality of gauges corresponding to the scores of each of the plurality of items, thereby enabling the subject to intuitively refer to the total score and the item scores.

[0126] It should be noted that the present embodiment is provided to facilitate understanding of the present invention and is not intended to limit the present invention. The present invention may be modified or improved without departing from the spirit thereof, and equivalents thereof are also included in the present invention.

[0127] In addition, in the present invention, a "unit" does not simply mean a physical means, but also includes cases where the functions of the "unit" are realized by software. Furthermore, the functions of one "unit" or device may be realized by two or more physical means, devices, or software, and the functions of two or more "units" or devices may be realized by one physical means, device, or software. [Explanation of symbols]

[0128] 100 score processing system, 110 memory unit, 115 acquisition unit, 120 setting unit, 123 data distribution generation unit, 125 regularity index data generation unit, 130 regularity score calculation unit, 135 item score calculation unit, 140 harmony score calculation unit, 145 total score calculation unit, 148 age prediction unit, 150 output unit, 155 display unit, 200 subject device

Claims

1. an event information acquisition unit that acquires event information of a predetermined event type related to the living situation of the subject, the event information being related to the occurrence timing of an event of the event type; a data distribution generating unit that generates a data distribution representing the occurrence timing for each predetermined time window on a predetermined time scale that is smaller than the predetermined time scale based on the event information; a regularity index data generation unit that generates regularity index data indicating regularity of the event type that can be expressed by the occurrence timing for each of the predetermined time windows based on the data distribution; an output unit that outputs the regularity index data; An information processing system comprising:

2. the predetermined event type includes a plurality of event types, the event information acquisition unit acquires the event information for each of the plurality of event types; the data distribution generation unit generates the data distribution for each of the plurality of event types; the regularity index data generating unit generates a plurality of individual regularity index data corresponding to each of the plurality of event types; The information processing system according to claim 1 .

3. The information processing system according to claim 2 , further comprising a display unit that displays a graphic based on each of the plurality of individual regularity index data for each of the predetermined time windows based on the plurality of individual regularity index data.

4. A setting unit that sets a time scale when generating the data distribution, the regularity index data generation unit generates the regularity index data on the set time scale. The information processing system according to any one of claims 1 to 3.

5. The information processing system according to any one of claims 1 to 3, wherein the regularity index data generation unit generates the regularity index data based on a similarity between contents of consecutive events on a predetermined time scale for each of the predetermined time windows.

6. The information processing system according to any one of claims 1 to 3, further comprising a regularity score calculation unit that calculates a regularity score indicating regularity of the event type based on the degree of deviation between a data distribution regarding the occurrence timing of the event information and a reference data distribution that is a reference data distribution.

7. 7. The information processing system according to claim 6, wherein the regularity score calculation unit calculates the regularity score according to at least one of KL divergence, Wasserstein distance, and circular Wasserstein distance between the data distribution and the reference data distribution.

8. 4. The information processing system according to claim 1, wherein the event type includes at least one of an activity status, a sleep status, and a nutritional status of the subject.

9. One or more processors included in the information processing device Acquiring event information of a predetermined event type related to the living situation of the subject, the event information being related to the occurrence timing of an event of the event type; generating a data distribution representing the occurrence timing for each predetermined time window on a predetermined time scale based on the event information; generating regularity index data indicating regularity of the event type that can be expressed by the occurrence timing for each of the predetermined time windows based on the data distribution; outputting the regularity index data; An information processing method that performs the above.

10. On the computer, Acquiring event information of a predetermined event type related to the living situation of the subject, the event information being related to the occurrence timing of an event of the event type; calculating a regularity index indicating regularity regarding the occurrence timing of the event information for each predetermined time window on a predetermined time scale, the predetermined time scale being smaller than the predetermined time scale, based on a data distribution regarding the occurrence timing based on the event information; outputting the regularity index; A program that executes the following.

Images